skill-optimize

Installation
SKILL.md

SKILL:OPTIMIZE_LLM_SKILL

Purpose: compress a verbose Markdown skill into terse, token-optimized form with zero logic loss.

[CONTEXT: Context window constraints require stripping grammar filler, structural bloat, and padding while preserving 100% logic.]

EXEC_FLOW

What it does: the ordered pipeline applied to every input skill.

flowchart LR
  A[1 PARSE] --> B[2 FRONTMATTER] --> C[3 STRIP] --> D[4 CONVERT] --> E[5 TAG] --> F[6 EMIT]
  1. PARSE: Read input Markdown. Isolate core objectives, variables, sequencing, and constraints.

  2. FRONTMATTER: Enforce the target skill's frontmatter to this template (keep the target's own name and description; never replace them with this skill's values):

Installs
3
Repository
dcoon/agent-kit
First Seen
Oct 26, 2026